Hushscript
Private transcription without the local AI setup.
A product I designed and built — browser-based audio and video transcription that turns a recording into an editable, speaker-labeled transcript. No local AI models, no subscription.
What it is
Hushscript turns an audio or video recording into a clean, speaker-labeled transcript you can edit, translate, analyze, and export. It runs entirely in the browser — no local AI models to install, no API keys, no command line.
I built it for people whose recordings are sensitive or valuable — journalists, lawyers, researchers, therapists, founders. Video never leaves your device, the audio sent for transcription is deleted after processing, and there is no subscription: you pay only for the minutes you use.
What it does
Speaker-labeled transcripts
Automatic speaker detection, included at no extra cost. Rename speakers and edit inline.
~99 languages
Transcribe in about 99 languages, detected automatically, with EU data residency available.
Free AI insights
One optional pass adds summaries, action items, decisions, chapters, and key quotes.
21 export formats
TXT, SRT, VTT, DOCX, PDF, JSON and more — plus password-protected archives. No watermark.
Privacy by design
Video stays on your device, transcription audio is deleted after processing, and stored data is encrypted.
No subscription
Prepaid minutes instead of a monthly plan. Start with 30 free minutes.
Links
Explore Hushscript
Follow Hushscript